Multivariate Analysis Of Prognostic Factors In 79 Patients With Glimoa And Correlations Of Some Factors

Objective: First to study the survival and prognostic factors of the patients with glioma as the theoretical reference to the clinical treatment, and to make an accuracy judgment of prognosis of the patients with glioma,therefore we can improve the survival status of the patients with glioma.And second to study the correlations of some factors.Methods: To review the data of 79 patients with glioma proved by surgery and pathology treated in the Second Hospital of Hebei Medical University between January 2005 and June 2009. Among the 79 patients (38 male and 41 female) 30 patients are oder then 50 years old and 49 patients are younger than 50 years old.There are 69 astrocytomas, 3 Glioblastomas(GBM), 4 oligodendrogliomas,1 mixed glioma,1 ependymoma and 1 glioma sarcomatosum. The medical records of the patients were reviewed for age, gender, the pre-operative neurological statusas defined by the Karnofsky Performance Status (KPS) , duration of symtoms , having epilepsy or not before operation, size of tumor, extent of resection, pathological grade, with or without postoperative radiotherapy and chemotherapy and the expression of PTEN, P53, PCNA, MGMT, MMP-9, P170 as possible prognostic factors. The date of operation was used as start points,while the date of death or last communication of the patient was used to establish end points. The extent of resection of tumor was based on operative record and postoperative CT and/or MRI.Statistical SPSS 17.0 software was used for uni- and multivariate cox regression. The survival data were first analyzed using the Kaplan-Meier method. Differences in survival between each group were tested for statistical significance by the log-rank test. After that, multivariate test method (Multiple Cox Regression) was performed to identify independent prognostic factors with P<0.05 in the results of univariate analysis.At last, linear relationship analysis with Spearman correlation coefficients was used to study the correlations of some factors.Results:Following a univariate regression analysis, good preoperative neurological status (higher KPS),longer duration of symtoms, having epilepsy before operation, macroscopically total resection,lower pathological grade, postoperative chemotherapy and lower expression of MMP-9 were asssciated with better overall survival rates.Multivariate cox regression analysis found that KPS, duration of symtoms, having epilepsy or not before operation, extent of resection, and postoperative chemotherapy were independent prognositic factors influencing surival.In linear relationship analysis, pathological grade was related to the expression of PCNA. In high pathological grade group, the positive rate of the expression of PCNA is higher than the low group. The correlation coefficient was 0.250(P=0.027).Secondly, the expression of PTEN was also related to the expression of P170,but the correlation coefficient was -0.252(P=0.025). At last, the expression of P170 was also related to the expression of MMP-9, and the correlation coefficient was 0.382(P=0.001).Conclusion:KPS≥70, having epilepsy before operation, macroscopically total resection and postoperative chemotherapy were the better independent prognostic factors.However,age, gender,duration of symtoms, size of tumor, pathological grade,postoperative radiotherapy and the expression of PTEN, P53, PCNA, MGMT, MMP-9, P170 were not revealed to be independent prognostic factors in our study.Therefore,acquiring the information of the patients with glioma,using the modern examnation and the effective therapy,we can improve the prognosis of the patients with glioma and prolong their lives.