Objective To investigate the effect of bufalin on chemosensitivity to human osteosarcoma cells Saos-2 and U2OS in vitro. Methods Saos-2 and U2OS cells were treated with bufalin at nontoxic dosage in combination with 0.01, 0.1, 1.0μg／ml doxorubicin（ADM） or 0.5, 1.0, 2.0μg／ml cisplatin（DDP）， cell proliferation was determined by CCK-8 assay， the morphological changes of cell apoptosis were observed by Hoechst 33258 staining， and apoptosis was assessed by flow cytometric analysis. Results CCK-8 assay showed that proliferation inhibition of bufalin on human osteosarcoma cells Saos-2 and U2OS was in a dose-dependent manner， and 0.005μmol／L bufalin in combination with ADM or DDP could enhance proliferation inhibition of Saos-2 and U2OS cells compared with ADM or DDP alone（q＞1.15）. Hoechst 33258 staining indicated that bufalin combined with ADM or DDP showed more obvious morphological changes of apoptosis in cells than ADM or DDP alone. Flow cytometric analysis showed that the apoptosis rates of 0.005μmol／L bufalin combined with 1.0μg／ml DDP or 2.0μg／ml ADM was significantly higher than those of cells treated with ADM or DDP alone（q＞1.15, P＜0.05）.
ConclusionBufalin at non-toxic dosage can enhance ADM or DDP-induced cell apoptosis of human osteosarcoma cells， showing effect of chemotherapy sensitization.

Objective To investigate the influence of angiotensin Ⅱ type 1 receptor blocker candesartan on growth and angiogenesis of mice bearing human hepatocellular carcinoma xenograft.Methods Fourty-eight mice bearing H22 hepatocellular carcinoma xenograft were established and randomly divided into normal saline group，low-dose candesartan group（2mg／kg），high-dose candesartan group（20mg／kg）and flurouracil group（25mg／kg）with 12 in each group. Nine days after administration， the mice were sacrificed， and the weight of transplanted tumors was measured to calculate the tumor inhibitory rate. The immunohistochemical technology was adopted to detect the expressions of vascular endothelial growth factor（VEGF） and CD34 to calculate microvessel density（MVD）. The same test was applied to 40 mice to observe the survival time and calculate the life extension rate. Results The tumor inhibition rates in highdose candesartan group（35.3％）and fluorouracil group（50.6％） were higher than 20.7％ in low-dose candesartan group（P＜0.05）. Compared with the normal saline group，significant reduction of VEGF expression was observed in lowdose（5.083±1.240） and high-dose candesartan（4.083±1.165） with significance（P＜0.05）. VEGF score in high-dose candesartan was lower than that in low-dose candesartan（P＜0.05）. Compared with the normal saline group， the MVD in low-dose（20.633±2.171） and high-dose candesartan groups（17.150±2.713） were significantly reduced（P＜0.01）. Compared with low-dose candesartan group， the MVD in highdose candesartan group were significantly reduced（P＜0.05）. The life extension rates in low-dose candesartan group（20.4％） was lower than 39.5％ in high-dose candesartan group and 30.6％ in fluorouracil group（P＜0.05）. Conclusion Candesartan is effective in inhibiting the growth of hepatocellular carcinoma xenograft， and the possible mechanism may be related to the inhibition of angiogenesis.

Objective To invistigate the discordance of epidermal growth factor receptor（EGFR） mutations between primary and brain metastatic tumors in non-small cell lung cancer（NSCLC）, and then analyze the consistency between them. Methods Thirty-one paired primary and brain metastatic tumors colleted from Jan. 2003 to Dec. 2011 were evaluated for the EGFR mutation by direct DNA sequencing and the cases of discordance of EGFR mutation were further detected by amplifyication refractory mutation system （ARMS）. Results In 31 case of primary tumors， 8 presented with deletions in exon 19 and 4 presented with L858R （leucine to arginine at codon 858） mutation in exon 21. In 31 case of brain metastatic tumors， 8 presented with deletions in exon 19 and 3 presented with L858R （leucine to arginine at codon 858） mutation in exon 21. There was no difference in distribution between the primary and brain metastatic tumors in NSCLC（P=0.793）.However， contrasted the primary and brain metastatic tumors individual through the ARMS， we found the EGFR gene mutation types were discordant in 5（16.1%） patients （4 male and 1 female）， including 3 adenocarcinoma （the brain metastatic tumors presented with deletions in exon 19 while the primary tumors presented with no EGFR gene mutation in 2 patients, and the primary tumor presented with L858R mutation in exon 21 while the brain metastatic tumor presented with no EGFR gene mutation in 1 patient）， 1 squamous-cell carcinoma and 1 atypical carcinoid （the primary tumors presented with deletions in exon 19 while the brain metastatic tumors presented with no EGFR gene mutation in both patients）. Conclusion EGFR mutations show discordance between the primary and the brain metastatic tumors in NSCLC.