RESUMO
Objective:To investigate the related factors of early postoperative recurrence of glioma patients and to establish a prediction model for early recurrence.Methods:A total of 94 patients with pathologically diagnosed glioma treated at Affiliated Hospital of Xuzhou Medical University from August 2014 to July 2016 were retrospectively analyzed. Kaplan-Meier method was used for survival analysis and log-rank test was carried out. Cox proportional risk regression model was used to analyze the clinical factors influencing early postoperative recurrence of glioma patients, and the prediction model of early recurrence was established.Results:The recurrence rates were 26.6% (25/94) and 39.4% (37/94) at 12 months and 24 months after operation, respectively. Univariate analysis showed that age ( χ2=9.59, P=0.008), degree of tumor resection ( χ2=14.26, P<0.001), Karnofsky performance status (KPS) score ( χ2=19.41, P<0.001), radiochemotherapy ( χ2=5.10, P=0.024) and pathological grade ( χ2=5.83, P=0.016) were significantly associated with early postoperative recurrence in glioma patients. Multivariate Cox proportional hazards regression model analysis showed that pathological grade ( OR=2.64, 95% CI: 1.75-3.97, P<0.001), degree of resection ( OR=0.34, 95% CI: 0.19-0.62, P<0.001) and radiochemotherapy ( OR=2.58, 95% CI: 1.34-4.99, P=0.005) were independent factors influencing early postoperative recurrence in glioma patients. The risk function model expression of early recurrence in glioma patients was h(t)=h 0exp(0.970X 1-1.081X 2+ 0.949X 3). X 1, X 2 and X 3 represented pathological grade, resection degree and radiochemotherapy respectively. Conclusion:High grade pathology and the absence of radiochemotherapy are independent predictors of early recurrence in glioma patients, and complete tumor resection can reduce the risk of early recurrence and improve the prognosis. The model of early recurrence prediction can provide some reference for clinical diagnosis and treatment.