探讨乳腺癌骨转移患者的临床病理特征，并分析其预后情况及相关影响因素。

根据纳入及排除标准，利用美国国立癌症研究所监测、流行病学和结果(SEER)数据库检索并筛选1975年1月至2016年12月5 815例转移性乳腺癌患者资料进行回顾性分析，评估了患者临床病理特征、治疗方式及其预后。其中，乳腺癌骨转移组3 146例，乳腺癌非骨转移组2 669例。按照预后情况，将3 146例乳腺癌骨转移患者分为2个亚组：死亡组1 669例和存活组1 477例。利用χ²检验和Mann-Whitney U检验比较骨转移和非骨转移组患者临床病理特征的差异；用二元Logistic回归分析乳腺癌骨转移的影响因素；用Kaplan-Meier法进行生存分析，并用单因素log-rank检验分析乳腺癌骨转移患者中死亡组与存活组临床病理特征的差异；用多因素Cox比例风险回归模型筛选影响乳腺癌骨转移者生存情况的独立因素。

骨转移组和非骨转移组患者在T分期、N分期、组织学分级、人种、ER、PR、HER-2、肿瘤分子分型和预后方面比较，差异均有统计学意义(Z＝-5.71、-2.39、-13.87、χ²＝14.55、305.74、245.56、69.34、335.36、79.15，P均<0.050)，2组间年龄、性别和原发灶位置比较，差异均无统计学意义(χ²＝0.57、2.71、0.45，P均>0.050)。Logistic回归分析结果显示：ER阳性、PR阳性、肿瘤T分期高和N分期高为导致乳腺癌患者骨转移的危险因素(OR＝1.775，95%CI：1.258～2.505，P＝0.001；OR＝1.425，95%CI: 1.236～1.643，P<0.001；OR＝1.095，95%CI：1.043～1.149，P<0.001；OR＝1.396，95%CI: 1.246～1.564，P<0.001)，而组织学分级越高，发生骨转移的风险反而越小(OR＝0.815，95%CI：0.733～0.907，P<0.001)。骨转移组与非骨转移组患者的OS比较，差异均具有统计学意义(χ²＝133.53，P<0.001)。骨转移患者中，2个亚组(死亡组和存活组)患者在T分期、N分期、组织学分级、年龄、ER、PR、HER-2、肿瘤分子分型、原发灶手术、放射治疗和化疗方面比较，差异均有统计学意义(Z＝-7.75、-3.22、-8.14、χ²＝39.80、69.81、87.45、51.87、132.47、36.24、6.05、36.24，P均<0.050)。Cox比例风险回归模型多因素分析结果显示：年龄、T分期、N分期、PR、HER-2、肿瘤分子分型、组织学分级、化疗、放射治疗和原发灶手术是影响骨转移组患者预后的独立因素(HR＝1.349，95%CI: 1.195～1.523，P<0.001；HR＝1.151，95%CI: 1.101～1.203，P<0.001；HR＝ 1.077，95%CI: 1.033～1.123，P<0.001；HR＝ 0.715，95%CI: 0.626～0.817，P<0.001；HR＝0.695，95%CI: 0.627～0.770，P<0.001；HR＝1.349，95%CI: 1.260～1.414，P<0.001；HR＝1.371，95%CI: 1.261～1.489，P<0.001；HR＝0.626，95%CI：0.562～0.697，P<0.001；HR＝0.874，95%CI：0.791～0.966，P＝0.008；HR＝0.663，95%CI: 0.561～0.784，P<0.001)。

乳腺癌骨转移患者预后优于非骨转移患者，与年龄、T分期、N分期、PR、HER-2、肿瘤分子分型、组织学分级有关，治疗方面原发灶手术、放射治疗和化疗有助于改善骨转移患者的预后。

To explore the clinicopathological characteristics of breast cancer patients with bone metastasis, and analyze their prognosis and related influencing factors.

According to the inclusion and exclusion criteria, totally 5 815 patients with metastatic breast cancer from January 1975 to December 2016 were screened out from the Surveillance, Epidemiology and Results (SEER) database of the U. S. National Cancer Institute for a retrospective analysis. The clinicopathological characteristics, treatment methods and prognosis of patients were evaluated. Among them, there were 3 146 patients with bone metastasis and 2 669 without. According to their prognosis, 3 146 patients with bone metastasis were divided into two subgroups: 1 669 patients in the death group and 1 477 patients in the survivor group. The χ² test and Mann-Whitney U test were used to compare the clinicopathological characteristics between bone metastasis group and non-bone metastasis group, binary Logistic regression was used to analyze the influencing factors of bone metastasis in breast cancer, the Kaplan-Meier method was used for survival analysis. The univariate log-rank test was used to analyze clinicopathological characteristics between the dead cases and survivors in 3 146 patients with bone metastasis, and the multivariate Cox proportional hazards regression model was used to find risk factors affecting the survival of patients.

There were significant differences in T stage, N stage, pathological grade, race, ER, PR, HER-2, molecular subtyping and prognosis between the bone metastasis group and non-bone metastasis group (Z=-5.71, -2.39, -13.87, χ²=14.55, 305.74, 245.56, 69.34, 335.36, 79.15, all P<0.050), while patient age, gender and position of primary tumor presented no significant difference (χ²=0.57, 2.71, 0.45, all P>0.050). The result of Logistic regression analysis showed that ER positive, PR positive, advanced T stage and advanced N stage were independent risk factors of bone metastasis in breast cancer patients(OR=1.775, 95%CI: 1.258-2.505, P=0.001; OR=1.425, 95%CI: 1.236-1.643, P<0.001; OR=1.095, 95%CI: 1.043-1.149, P<0.001; OR=1.396, 95%CI: 1.246-1.564, P<0.001), while advanced pathological stage indicated lower risk of bone metastasis (OR=0.815, 95%CI: 0.733-0.907, P<0.001). The overall survival showed a significant difference between the bone metastasis group and non-bone metastasis group (χ²=133.53, P<0.001). In patients with bone metastasis, there were significant differences in T stage, N stage, pathological grade, patient age, ER, PR, HER-2, molecular subtyping, primary tumor surgery, chemotherapy and radiation therapy between death group and survivor group (Z=-7.75, -3.22, -8.14, χ²=39.80, 69.81, 87.45, 51.87, 132.47, 36.24, 6.05, 36.24, all P<0.050). The multivariate analysis of Cox proportional hazards model showed that patient age, T stage, N stage, PR, HER-2, molecular subtyping, pathological grade, chemotherapy, radiotherapy and primary tumor surgery were independent factors affecting the prognosis of breast cancer patients with bone metastasis(HR =1.349, 95%CI: 1.195-1.523, P<0.001; HR=1.151, 95%CI: 1.101-1.203, P<0.001; HR=1.077, 95%CI: 1.033-1.123, P<0.001; HR= 0.715, 95%CI: 0.626-0.817, P<0.001; HR=0.695, 95%CI: 0.627-0.770, P<0.001; HR=1.349, 95%CI: 1.260-1.414, P<0.001; HR=1.371, 95%CI: 1.261-1.489, P<0.001; HR=0.626, 95%CI: 0.562-0.697, P<0.001; HR=0.874, 95%CI: 0.791-0.966, P=0.008; HR=0.663, 95%CI: 0.561-0.784, P<0.001).

The prognosis of breast cancer patients with bone metastasis is superior to that of the patients with non-bone metastasis, related to patient age, T stage, N stage, PR, HER-2, molecular subtyping and pathological grade. Primary tumor surgery, radiotherapy and chemotherapy can significantly improve the prognosis of patients with bone metastasis.