中文摘要：

目的对四种乳腺癌骨转移动物模型的构建方法进行比较研究。方法32只4～6周龄雌性裸鼠随机分为A、B、C、D4组，每组8只，每只裸鼠以5×10^5个MDA．MB．231细胞按分组分别注射人左第二乳房脂肪垫、尾静脉、左心室和左胫骨骨髓腔。观察A组裸鼠乳腺原位成瘤情况、全组注射致死情况和全组49d后骨转移发生情况。结果A组乳腺原位移植瘤形成率为87．5％（7／8）。仅C组注射致死裸鼠1只。各组骨转移发生率分别为：0（0／8）、12．5％（1／8）、71．4％（5／7）、100％（8／8）。A组和C组、A组和D组、B组和C组、B组和D组之间骨转移发生率差异均有统计学意义。结论肿瘤细胞经乳腺原位移植和尾静脉注射移植两种方法不适用于建立乳腺癌骨转移模型；肿瘤细胞经左心室注射移植和骨原位移植方法能稳定地获得骨转移，是乳腺癌骨转移动物模型构建的可行方法。

英文摘要：

Objective To investigate four methods for establishing animal models of human breast cancer bone metastasis, Methods Thirty-two female nude mice aged 4-6 weeks were divided randomly into four groups （n=8 in each group）. 5xlO5 MDA-MB-231 cells were injected into the body via the left second mammary fat pads （group A）, the tail veins （group B）, the left heart ventricles （group C） and the left tibia marrow cavities （group D）, respectively. Tumor formations in situ were recorded in group A. Deaths after the injection were recorded. The surviving nude mice 49 days after the injection were subjected to pathological examination to determine bone metastasis. Results The rate of tumor formation in situ of group A was 87.5 %（7/8）. One mouse in group C died after the injection of MDA-MB-231 cells. The bone metastasis rate in groups A, B, C and D was zero （0/8）, 12.5 % （1/8）, 71.4 % （5/7） and 100 % （8/8）, respectively. There was statistically significant difference in the bone metastasis rate between group A and group C, group A and group D, group B and group C; and group B and group D. Conclusion Injections of tumor cells via the breast fat pads and tail veins were not suitablemethods to establish animal models of human breast cancer bone metastasis. The bone metastasis model could be established efficiently by injecting tumor ceils into the left heart ventricles or the bone marrow cavity of nude mice.