Multivariate Analysis For Prognostic Factors Of Osteosarcoma: A Study In 43 Cases

Objective:To explore the prognostic factors of osteosarcoma for survival.Methods:the clinical data of 43 patients with osteosarcoma between 2005.1～2007.3 were retrospectively reviewed. Patient's sex, age, tumor site, course of disease, serum alkaline phosphatase (AKP) level, preoperative chemotherapy, Enneking surgical stage, surgical method, distant metastasis were analyzed by Kaplan-Meier method, Log-rank test and COX regression model.The numeration data were analyzed byχ2 or Fisher exact test.Results:43 patients had at least 3 years follow up,28 patients were alive, while 15 patients died. Live time was between 6 to 65 months with a median survival of 42 months. Overall survival rate of three years was 65.1%. Univariate analysis revealed that the prognosis of osteosarcoma was significantly related to tumor site (P=0.01), Enneking surgical stage (P=0.002), surgical method (P<0.0001) and distant metastasis (P=0.002). Multivariate analysis by COX regression model suggested Enneking surgical stage (P=0.028), surgical method (P=0.001) and distant metastasis (P=0.007) were the independent prognostic factors. The numeration data were analyzed by Fisher exact test. Though the preoperative chemotherapy was unrelated with the survival of osteosarcoma, we found the sensitivity to preoperative chemotherapy was an important factor that affected the prognosis of osteosarcoma (P=0.007). Furthermore, we found AKP level after preoperative chemotherapy was a good method to evaluate the effect of neoadjuvant chemotherapy.Conclusion:This trial indicated that the prognosis of osteosarcoma was significantly correlated with Enneking surgical stage, surgical method and distant metastasis, early detection and wide excision were interventional methods to improve the survival of osteosarcoma. Interestingly, AKP level after preoperative chemotherapy may be a good method to evaluate the effect of neoadjuvant chemotherapy.