Clinicopathologic Features Of Nonmetastatic Osteosarcoma And Clinical Predication Model

Objective:The research project was conducted to probe into clinicopathologic features of nonmetastatic osteosarcoma and relevant risk factors for pulmonary metastasis in the long run,establish clinical predication models for nonmetastatic osteosarcoma and open up referential idea to clinical treatment.Methods: Schemes were conceived based on experiments and U.S.SEER data was reviewed.Totally 1,490 osteosarcoma patients confirmed between 2004 and 2016 were included.They were divided into nonmetastatic group(1,379 cases)and pulmonary metastasis group(111 cases),depending on whether there was pulmonary metastasis to the patients.The optimal cut-off values for age while confirmation of the disease,tumor size and other continuous variables were finalized.Such continuous variables were then turned into categorical variables in line with the finalized optimal cut-off values.Successively values were assigned to relevant clinicopathologic features.Comparisons of the differences in relevant pathological features of the two groups of patients through chi-square tests.Factors with statistical significance during chi-square tests were included into the binary logistic regression mode.l.Relevant risk factors for pulmonary metastasis to osteosarcoma patients in the long run were made clear.ROC curve was drawn to evaluate the predication effects of the model.2.Analysis of prognostic factors for nonmetastatic osteosarcoma group(1,379 cases).The patients with nonmetastatic osteosarcoma were randomly divided into model group(965 cases)and verification group(414 cases)by SPSS software.Kaplan-Meier was adopted to compute the overall survival rate and cancer specificity survival rate of the model group,and survival curves were drawn.The differences in factors related to single factor were compared through Log-Rank test.Then variables with statistical significance from the single-factor analysis were included into Cox proportional hazard model,and independent factors affecting prognosis of patients with nonmetastatic osteosarcoma were ascertained.3 Evaluation of the predication model.The nomogram for nonmetastatic osteosarcoma patients was established based on the results of the Cox proportional hazard model.By way of internal data verification,C index was computed to evaluate the model predication accuracy.Results: patients with nonmetastatic osteosarcoma had favorable prognosis.The 5-year and 10-year overall survival rate was 70.4% and63.4% respectively,and the 5-year and 10-year cancer specificity survival rate was 71.8% and 66.1%.2 According to the chi-square test results,the differences,between the nonmetastatic osteosarcoma group and the pulmonary metastasis group,in age while confirmation,T-staging(AJCC 6th),primary site of the tumor,pathological patterns,tumor differentiation degree and other clinicopathologic features were statistically significant(P<0.05).The binary logistic regression model suggested that age while confirmation and tumor size came as independent risk factors for pulmonary metastasis of osteosarcoma patients.The area under the ROC curve was 0.71,95% CI(0.664～0.755),with desirable predication effects.3 According to single-and-multi-factor prognosis analysis,age while confirmation,T-staging(AJCC 6th),primary site of the tumor,tumor size,tumor differentiation degree,surgical situation,pathological patterns,marital status and other clinicopathologic features came as independent factors for prognosis of nonmetastatic osteosarcoma patients(P<0.05).For the model group,the C indexes of the overall survival rate and cancer specificity rate were0.736 [95% CI(0.705,0.767)] and 0.745 [95%,CI(0.714,0.776)]respectively.For the verification group,the C indexes of the overall survival rate and cancer specificity rate were 0.67 [95% C(I 0.617,0.723)] and 0.657 [95% CI(0.603,0.711)] respectively.The model performed well in predication.Conclusions: 1 Nonmetastatic osteosarcoma has favorable prognosis.Age while confirmation and tumor size come as relatively risk factors for pulmonary metastasis of nonmetastatic osteosarcoma in the long run.Much importance should be attached to follow-up visits and regular physical examinations.2 Age while confirmation,T-staging(AJCC 6th),primary site of the tumor,tumor size,tumor differentiation degree,surgical situation,pathological patterns and marital status come as independent factors affecting the prognosis of nonmetastatic osteosarcoma.3 Clinical predication model can provide a qiuck and convenient individual assessment of prognosis.