目的: 评价卡铂联合紫杉醇密集化疗新辅助治疗三阴性乳腺癌的疗效和远期生存，探索三阴性乳腺癌新辅助化疗的优化方案。

方法: 我院病理确诊的三阴性乳腺癌（临床分期T1-4/N0-3M0）,接受密集方案（紫杉醇175 mg/m² d1+卡铂AUC 4 d1，q14d）新辅助化疗后手术的患者，采用倾向性评分匹配（PSM）按1:1与接受3周方案（紫杉醇175 mg/m² d1+卡铂AUC 5 d2，q21d）新辅助治疗后手术的患者进行最近邻匹配。主要研究终点：两组患者的病理学完全缓解率（pCR，ypT0/is ypN0）；次要研究终点：客观缓解率（ORR）、无复发生存（RFS）、总生存(OS)、安全性。

结果:  2008年1月-2018年9月纳入IIA-IIIC期三阴性乳腺癌患者100例，中位年龄47岁（24-75岁），绝经前65例。密集组和3周组各50例。密集组中49例完成4-6周期化疗，1例完成3周期化疗。3周组中42例完成4-6周期化疗，8例完成2-3周期化疗。100例患者均可评价临床疗效，完全缓解（CR）11例，部分缓解（PR）79例，ORR为90%，其中密集组和3周组ORR均为90%。100例患者完成新辅助化疗后均接受了手术治疗，pCR率为36%（36/100），其中密集组和3周组pCR率相似，分别为34%（17/50）和38%（19/50），P=0.835。18例检测BRCA1/2基因的患者中，4例BRCA1/2基因突变，其中2例达pCR。化疗最常见的不良反应为中性粒细胞减少，由于预防性重组人粒细胞集落刺激因子的应用，密集组3/4级中性粒细胞减少的发生率低于3周组（分别为32.7% vs. 68.0%，P=0.001），但肝功能异常的发生率更高（分别为57.1% vs. 32%，P=0.015）。中位随访55个月（3~150个月），全组5年RFS为77.3%，其中密集组5年RFS（83.5%）好于3周组（75.2%），但无统计学差异（P=0.809）；全组5年OS为85.7%，密集组与3周组的OS相似，分别为87.9%、84.5%，P=0.624。全组获得pCR的患者与非pCR者相比5年RFS（90.7% vs. 69％，P=0.032）和OS（100% vs. 76.9%，P=0.002）显著提高。

结论: 卡铂联合紫杉醇密集方案与常规3周方案疗效相似，耐受性良好；鉴于其能缩短治疗时间，是三阴性乳腺癌新辅助化疗的可选方案。

Objectives: To compare the efficacy of dose-dense (biweekly) carboplatin plus paclitaxel (PC) with standard 3-weekly PC as neoadjuvant chemotherapy (NAC) in triple-negative breast cancer (TNBC), and to explore an optimal neoadjuvant chemotherapy regimen for TNBC.

Methods: Patients with core needle biopsy confirmed pathological diagnosis of TNBC who received dose-dense PC (paclitaxel 175 mg/m² d1, carboplatin AUC 4 d1，every 2 weeks) as NAC were 1:1 matched using propensity score matching (PSM) with TNBC patients who received standard 3-weekly PC (paclitaxel 175 mg/m² d1, carboplatin AUC 5 d2，every 3 weeks).The primary endpoint was rate of pathologic complete response (pCR, ypT0/is ypN0). The secondary endpoint were objective response rate(ORR)、 recurrence-free survival(RFS)、overall survival(OS) and safety.

Results: Between January, 2008 and September, 2018, 100 TNBC patients of stage IIA-IIIC were enrolled (dose-dense group 50 patients, standard group 50 patients) with a median age of 47 years (range 24-75 years). 65 patients were premenopausal. In dose-dense group, 49 patients completed 4-6 cycles of NAC, and 1 patient completed 3 cycles. In standard group, 42 patients completed 4-6 cycles of NAC, and 8 patients completed 2-3 cycles. Of the 100 patients who were evaluated for clinical response, 11 patients had complete response(CR), and 79 patients had partial response(PR), and the overall ORR was 90%. The ORR of dose-dense group and standard group were the same (90%). All patients underwent surgery after NAC and the overall pCR rate was 36%(36/100). The pCR rates were similar between the two groups: 34% (17/50) in dose-dense group and 38% (19/50) in standard group, respectively (P=0.835). 18 patients assessed BRCA1/2 mutation status, and 4 patients were BRCA1/2 mutation carriers, of which 2 patients had  pCR. The most common adverse events were neutropenia. Due to the use of granulocyte colony stimulating factor (G-CSF), grade 3/4 neutropenia was less in dose-dense group than in standard group (32.7% vs. 68.0%，P=0.001), while the rate of ALT/AST elevation was higher(57.1% vs. 32.0%，P=0.015) in dose-dense group. The median follow-up time was 55 months (3~150 months). In overall population, 5-year RFS was 77.3%. 5-year RFS was non-significantly higher in dose-dense group than in standard group (83.5% vs.75.2%，P=0.809). 5-year OS was 85.7% in overall population, and similar in dose-dense and standard group(87.9% vs. 84.5%，P=0.624). Patients who achieved pCR after NAC had significantly better 5-year RFS(90.7% vs. 69%，P=0.032) and 5-year OS(100% vs. 76.9%，P=0.002)than patients with residual disease.

Conclusions: Dose-dense PC had similar efficacy with standard 3-weekly PC and had a good safety profile. Since dose-dense regimen could shorten the duration of therapy, it could be an alternative in TNBC.