探讨乳腺癌患者长期的OS和无复发生存(RFS)及其影响因素。

依据纳入、排除标准，选取2000年1月1日至2015年12月31日在解放军总医院第五医学中心乳腺外科治疗的乳腺癌患者作为研究对象进行回顾性分析，最终共有2 423例患者纳入随访。采用Kaplan-Meier法及Log-rank检验比较不同临床分期、不同分子分型以及不同手术方式的患者5年OS和RFS的差异。采用Cox比例风险回归模型分析患者生存的影响因素。

2 423例乳腺癌患者的中位年龄为48岁，随访时间为3.5～18.7年，中位随访时间为5.2年，随访率为85.3%(2 066/2 423)。本组患者5年、10年OS率分别为91.5%和84.4%，5年、10年RFS率(除外Ⅳ期乳腺癌)分别为85.8%和78.4%。在临床分期方面：本组可手术乳腺癌患者占81.0%(1 963/2 423)，局部晚期乳腺癌患者占15.6%(378/2 423)，Ⅳ期患者占3.4%(82/2 423)。0、Ⅰ、Ⅱ、Ⅲ、Ⅳ期患者5年OS率分别为100%、98.5%、93.8%、78.1%和50.8%，5年RFS率(除外Ⅳ期乳腺癌)分别为98.5%、95.4%、87.0%和63.0%。临床分期与患者5年OS及RFS明显关联(χ²＝356.067、250.433，P均<0.001)。在分子分型方面：排除205例分子分型不详者后，将剩余2 218例按照受体状况分为luminal型、HER-2过表达型和三阴性乳腺癌(TNBC)，3组的比例分别为72.3%(1 604/2 218)、10.1%(225/2 218)、17.5%(389/2 218)，5年OS率分别为93.1%、88.3%和84.4%，5年RFS率(除外Ⅳ期乳腺癌)分别为87.7%、84.8%和76.7%。3组患者间5年OS率及RFS率相比，差异均有统计学意义(χ²＝24.124、31.668，P均<0.001)。在手术方式方面：本组患者保留乳房(简称保乳)率为24.8%(600/2 423)。Ⅰ期患者保乳率为44.9%(309/688)，接受保乳者5年OS率比乳房全切者高(99.3%比98.4%，χ²＝6.338，P＝0.012)，但5年RFS率与乳房全切者相比，差异无统计学意义(96.7%比94.8%，χ²＝2.245，P＝0.134)；Ⅱ期患者保乳率为21.1%(237/1 125)，保乳手术者与乳房全切者5年OS率分别为97.2%和92.7%，5年RFS率分别为88.5%和87.1%，组间比较，差异均无统计学意义(χ²＝3.793、1.425，P＝0.051、0.233)。在患者生存、复发影响因素方面：Cox比例风险回归模型显示，患者的年龄、临床分期、分子分型均与OS有关(HR＝1.017，95%CI：1.004～1.029，P＝0.019; HR＝3.242，95%CI：2.763～3.803，P<0.001；HR＝1.203，95%CI：1.066～1.357，P＝0.003)。

年龄、临床分期、分子分型是乳腺癌患者预后的影响因素。对于Ⅰ、Ⅱ期乳腺癌患者而言，保乳手术的疗效优于或不差于乳房全切术。

To explore the long-term OS and recurrence-free survival (RFS) and the influencing factors in breast cancer patients.

According to the inclusion and exclusion criteria, a total of 2 423 breast cancer patients who underwent breast surgery in the Department of Breast Surgery, Fifth Medical Center, General Hospital of PLA from January 1, 2000 to December 31, 2015 were enrolled for a retrospective study. Kaplan-Meier method and log-rank test were used to compare the 5-year OS and RFS among patients with different clinical stages, molecular subtypes and surgical methods. Cox proportional hazard regression model was used to analyze the influencing factors on the patient survival.

In 2 423 cases of breast cancer, the median age was 48 years, the median follow-up time was 5.2 years (range: 3.5-18.7 years) and the follow-up rate was 85.3%(2 066/2 423). The 5-year and 10-year OS rate was 91.5% and 84.4% respectively, and the 5-year and 10-year RFS rate (except stage Ⅳ breast cancer) was 85.8% and 78.4% respectively. As for clinical stage, the operable breast cancer accounted for 81.0% (1 963/2 423), local advanced breast cancer 15.6%(378/2 423)and stage Ⅳ breast cancer 3.4%(82/2 423). The 5-year OS rate in stage 0, Ⅰ, Ⅱ, Ⅲ and Ⅳ breast cancer was 100%, 98.5%, 93.8%, 78.1% and 50.8%, respectively, and the 5-year RFS rate (except stage Ⅳ breast cancer) was 98.5%, 95.4%, 87.0% and 63.0%, respectively. Clinical stage was significantly correlated with the 5-year OS and RFS rate of patients (χ²=356.067, 250.433, both P<0.001). As for molecular subtype, except 205 patients with unknown molecular type, the remaining 2 218 patients were divided into luminal, HER-2 overexpression and triple-negative subtypes according to the receptor status, accounting for 72.3%(1 604/2 218), 10.1%(225/2 218) and 17.5%(389/2 218), respectively. The 5-year OS rate of luminal, HER-2 overexpression and triple-negative subtypes was 93.1%, 88.3% and 84.4%, respectively, and the 5-year RFS rate (except stage Ⅳ breast cancer) were 87.7%, 84.8% and 76.7%, respectively. The 5-year OS and RFS rate presented a significant difference across three groups (χ²=24.124, 31.668, both P<0.001). As for surgical methods, the breast-conserving rate was 24.8% (600/2 423) in all patients. In stage Ⅰ breast cancer, the breast-conserving rate was 44.9%(309/688); the breast-conserving surgery group had a higher 5-year OS rate compared with mastectomy group (99.3% vs 98.4%, χ²=6.338, P=0.012); however the 5-year RFS rate showed no significant difference between two groups (96.7% vs 94.8%, χ²=2.245, P=0.134). In stage Ⅱ breast cancer, the breast-conserving rate was 21.1%(237/1 125), the 5-year OS rate in breast-conserving surgery group and mastectomy group was 97.2% and 92.7%, respectively and the 5-year RFS rate was 88.5% and 87.1%, respectively, indicating no significant difference between two groups (χ²=3.793, 1.425; P=0.051, 0.233). Cox proportional hazard regression analysis showed that age, clinical stage and molecular subtype were correlated with the OS of breast cancer patients (HR=1.017, 95%CI: 1.004-1.029, P=0.019; HR=3.242, 95%CI: 2.763-3.803, P<0.001, HR=1.203, 95%CI: 1.066-1.357, P=0.003).

Age, clinical stage and molecular subtypes are prognostic factors for breast cancer. In stage Ⅰ-Ⅱ breast cancer, the efficacy of breast-conserving surgery is superior to or not worse than mastectomy.