摘要目的:探讨高迁移率族蛋白B1（HMGB1）蛋白在乳腺癌中的表达及临床病理意义。方法:收集2006—2018年浙江省东阳市人民医院病理科浸润性乳腺癌石蜡标本417例和同时切除的正常乳腺组织石蜡标本26例。采用免疫组织化学EnVision法检测和比较浸润性乳腺癌组织和正常乳腺组织中HMGB1蛋白的表达。分析HMGB1蛋白胞核高表达及胞质阳性表达与乳腺癌患者临床病理特征的关系。结果:（1）乳腺癌中HMGB1蛋白胞核高表达率及胞质阳性率分别为80.8%（337/417）和16.8%（70/417），而其在正常乳腺组织中分别为46.2%（12/26）和0。乳腺癌中HMGB1蛋白胞核高表达率和胞质阳性率均明显高于正常乳腺组织（分别 P<0.001， P=0.046）。（2）组织学级别高、雌激素受体（ER）阴性、孕激素受体（PR）阴性的乳腺癌患者，其HMGB1蛋白胞核高表达率明显更高（分别 P=0.006， P=0.004， P<0.001），胞质阳性率也明显更高（均 P<0.001）。Logistic回归模型多因素分析显示乳腺癌患者HMGB1蛋白胞核高表达的独立相关因素为肿瘤组织学分级（ OR=2.188，95 %CI=1.078~4.443， P=0.030），而HMGB1蛋白胞质阳性表达的独立相关因素包括肿瘤组织学分级（ OR=3.031，95 %CI=1.600~5.742， P=0.001）、ER（ OR=0.129，95 %CI=0.034~0.494， P=0.003）及TNM分期（ OR=3.820，95 %CI=1.042~14.001， P=0.043）。（3）Cox比例风险模型多因素分析显示HMGB1蛋白胞核高表达是影响乳腺癌患者总生存的独立危险因素（ HR=0.366，95 %CI=0.138~0.972， P=0.044）。 结论:HMGB1蛋白胞核高表达及胞质阳性表达与乳腺癌患者多项预后不良因素密切相关，有望成为抗乳腺癌治疗的一个潜在生物学标志。

abstractsObjective:To investigate the expression and clinicopathological significance of high mobility group box protein B1 (HMGB1) protein in breast cancer.Methods:The expression of HMGB1 protein in 26 normal breast tissues and 417 invasive breast cancer tissues diagnosed at Dongyang People′s Hospital, Zhejiang Province from 2016 to 2018 were detected by immunohistochemical EnVision method. The relationship between nuclear and cytoplasmic HMGB1 protein expression and clinicopathologic features of breast cancer patients were analyzed.Results:The nuclear and cytoplasmic expression of HMGB1 protein was 80.8% (337/417) and 16.8% (70/417) respectively in breast cancer, and was 46.2%(12/26) and 0(0/26) respectively in normal breast tissue. Both nuclear and cytoplasmic expression of HMGB1 protein in breast cancer were significantly higher than normal breast tissue ( P<0.001, P=0.046, respectively). The nuclear expression of HMGB1 protein was also higher in high grade, estrogen receptor (ER) negative, progesterone receptor (PR) negative ( P=0.006, P=0.004, P<0.001, respectively); whereas the cytoplasmic expression of HMGB1 protein was also higher in high grade, estrogen receptor (ER) negative, progesterone receptor (PR) negative ( P<0.001 in all) breast cancers. Multivariate logistic regression model showed that nuclear HMGB1 expression correlated with histologic grade ( OR=2.188, 95 %CI=1.078-4.443, P=0.030), while cytoplasmic HMGB1 expression correlated with histologic grade ( OR=3.031, 95 %CI=1.600-5.742, P=0.001), ER ( OR=0.129, 95 %CI=0.034-0.494, P=0.003) and TNM staging ( OR=3.820, 95 %CI=1.042-14.001, P=0.043). Multivariate analysis of Cox proportional hazard model showed that nuclear HMGB1 expression was an independent risk factor for the overall survival of breast cancer patients ( HR=0.366, 95 %CI=0.138-0.972, P=0.044). Conclusion:Nuclear and cytoplasmic HMGB1 proteins are related to multiple poor prognostic factors in breast cancer, and may be a potential biomarker for breast cancer treatment.